Power-Efficient Gathering in Sensor Information Systems (PEGASIS)
This hierarchical routing protocol is chain-based and aimed at promoting power efficiency. Its structure is based on that of LEACH, but it is also a LEACH’s improvement. Instead of creating clusters, PEGASIS creates chains that connect the sensor nodes, allowing for data transmission from each node to the neighboring node (Kour, 2012). One of the nodes in the chain is selected to send data to the base station. Any collected data moves from one node to the other, and it is aggregated in the process before it is eventually transmitted to the base station. The construction of the chain is constructed in a greedy way. One of the outstanding features of this routing protocol as compared to other protocols including the LEACH protocol is that it eliminates the dynamic cluster formation overhead, thus minimizing the total distances that need to be transmitted by the non-leader nodes and reducing the total number of transmissions, an aspect that makes it applicable in large-scale networks. Nevertheless, there is an excess of delay introduced by this protocol for the chain’s distant node. This taints its applicability in large-scale environments with unreliability, as such environments, including border monitoring, require constant and timely updates on the situations in order to facilitate timely response to any situations. In addition, such a system that only allows for a single leader may lead to such a leader becoming a bottleneck, thus further slowing the flow of information within the network and delaying responses to any situations that require rapid communication within borders (Singh, Singh, & Singh, 2010).
Location-based Routing Protocols
These protocols replace the node identifiers with areas as packet targets. As such, areas are identified an all the nodes within a certain area are considered acceptable destinations, with the capability of receiving and processing data within the network (Kour, 2012). Such routing stands out to be of great importance in sensor network contexts, especially in the borrowing of sensor data from a different region. In addition, information about the location can be helpful in promoting energy efficient data routing, as sensor networks are spatially deployed in a region and they do not have an addressing scheme such as IP-addresses (Jiang & Manivannan, 2004). This feature would allow such a routing protocol to succeed in a large-scale environment including a border. For instance in the case where a region requiring sensing is known, the location’s sensor nodes can be used in a way that will allow for the query to be directed to that specific region, and thus the number of transmissions will be eliminated significantly. These routing protocols highly consider the sensor nodes’ mobility, and perform extremely well with an increase in the network density. However, it is important to note that these protocols are very poor in performance in cases where there is sparse network deployment and data aggregation and extensive processing by the node that is heading the network is not promoted (Singh, Singh, & Singh, 2010). This seems to be the case on borders as they involve little activity, and thus fewer nodes are deployed on them for different purposes. As such, the likelihood of these routing protocols to succeed in border environments is minimal.
Gear and Energy Aware Routing protocol (GEAR)
This protocol utilizes neighbor selection heuristics that are geographically aware and energy aware in routing a packet that targets a given region. A learning cost and an estimated cost of getting to the destination via the neighbors is kept by each node in the GEAR routing (Singh, Singh, & Singh, 2010). In order to account for the routing that occurs around holes within the network, the estimated cost is refined into the learned cost. A hole occurs in the case where there is no neighboring node between a certain sensor node and its target region. There is equilibrium between the learned cost and the estimated cost in the case where there is no hole. Each time a packet reaches the target region, the learned cost is moved one hop back to allow for the adjustment of the next packets route setup. This algorithm involves two phases, starting with the movement of packets towards the selected region, and forward movement of packets within the selected region. When GEAR is compared to earlier geographical routing protocols such as GPRS, it is evident that the former is more energy efficient and efficient in delivery of packets. As such, this routing protocol may provide a considerable alternative in large-scale environments such as border monitoring.
Medium Access Control (MAC) Protocols
MAC protocols are design with various attributes that allow them to succeed in terms of information sharing in the WSNs. Energy efficiency is one of the most important attributes that is presented in the designs of these protocols (Li & Thai, 2008). Energy efficiency allows for a prolonged lifetime of the network and thus reduces disturbances in the data sharing process. Changes in the topology, network and node density of the network are inevitable. As such, the adaptability and scalability of the Mac protocols form other protocols that would allow for effective and rapid adaptation of the network to such changes. These changes may be because of new nodes addition to the network, reduced lifetime of the nodes, and a variety of disturbances that may interfere with the connectivity of the network and the entire topology. These primary attributes would allow the MAC protocol to adapt to such important changes. Nevertheless, latency, bandwidth utilization, and throughput form other attributes that may be secondary but still important to the success of the MAC protocol (Li & Thai, 2008).
MAC Layer Protocols
Sensor-MAC (S-MAC) Protocol
The S-MAC –protocol involve a system of timely sleep-listen schedules that are based on synchronizations that are locally managed (W. Ye & Heidemann, 2004). As such, virtual clusters are created by neighboring nodes in the creation of a shared sleep schedule. In the case where two neighboring nodes are from different virtual clusters, they are bound to wake up during the listen periods of the clusters they represent. However, there is a likelihood of sharing two distinct schedules, leading to more consumption of energy through overhearing and idle listening. Periodical packets of SYNC broadcasts facilitate the exchange of schedules to the immediate neighbors of nodes. As such, a SYNC packet is send by each node during the synchronization period. In order to avoid collision during communication, a carrier sense is applied. In the case of uncast forms of data packets, CTS/RTS packet exchanges are used for collision avoidance. One of the important features of S-MAC that may be of importance in the surveillance of borders as a form of large scale application is the message passing concept, which involves the division of long messages into frames that are sent in a burst(W. Ye & Heidemann, 2004). As a result of such a mode of communication, the communication overhead is reduced, thus saving energy at the expense of medium access unfairness.
Due to periodic sleeping, high latency may be experienced especially in the case of multi-hop routing algorithms as every immediate note has a different sleep schedule. Such a latency resulting from periodic sleeping is referred to as ‘sleep delay’ (W. Ye & Heidemann, 2004). This delay is a downside of the S-MAC protocol as it interferes with the rapid sharing of data. As such, it may lead to late communication in large-scale environments and inadequate response to situations. Sleep delays and the general latency can be improved through application of the adaptive listening technique. As such, this technique allows a node to wake up when it overhears the transmission of the neighbor, for a short period towards the finish of the transmission. Thus, the neighbor can immediately pass data in the case where the next-hop node is the awakened node. The energy efficiency of this protocol, facilitated by the sleep schedules, is an advantageous feature, as it would prolong the life of the protocol especially in large-scale environments that require less maintenance (W. Ye & Heidemann, 2004). In addition, sleep schedule announcements can be utilized in the prevention of synchronization overhead of time. Nevertheless, the probability of collision is increased by the lack of usability of CTS/RTS by the broadcast data packets. On the other hand, energy efficiency may be compromised by adaptive listening as it may lead to idle listening or overhearing in the case where the packet was not intended for the listening node. This aspect would lead to massive communication of unintended information in large-scale environments and waste of energy that would reduce the sustainability of the network and increase the cost of maintenance. In addition, border monitoring involves variable traffic load as the occurrence of events is not predictable, thus, the constant and predefined listen and sleep periods of this protocol make it less efficient in such circumstances (W. Ye & Heidemann, 2004).